Extinction vulnerability of coral reef fishes

Abstract

With rapidly increasing rates of contemporary extinction, predicting extinction vulnerability and identifying how multiple stressors drive non-random species loss have become key challenges in ecology. These assessments are crucial for avoiding the loss of key functional groups that sustain ecosystem processes and services. We developed a novel predictive framework of species extinction vulnerability and applied it to coral reef fishes. Although relatively few coral reef fishes are at risk of global extinction from climate disturbances, a negative convex relationship between fish species locally vulnerable to climate change vs. fisheries exploitation indicates that the entire community is vulnerable on the many reefs where both stressors co-occur. Fishes involved in maintaining key ecosystem functions are more at risk from fishing than climate disturbances. This finding is encouraging as local and regional commitment to fisheries management action can maintain reef ecosystem functions pending progress towards the more complex global problem of stabilizing the climate.

Figure 1

Predictive framework for assessing extinction risk. (a) The vulnerability of coral reef fish species populations to climate change disturbances (coral bleaching and mortality) against the risk that population decline is catastrophic and leads to species extinction. The vertical axis is an expert-weighted composite index including information on species specialization on coral and reef habitat for (1) diet, (2) habitat use and (3) settlement, and also including information on (4) species body size. The horizontal axis is an expert-weighted composite index of extinction risk and includes information on (1) geographical range size, (2) depth range, (3) occupancy and (4) numerical rarity. (b) Species data applied to the framework for 134 species of coral reef fishes in eight functional groups.

Figure 2

Assessing the climate vulnerability index through a major disturbance event. Predicted climate change vulnerability scores and fish species population change in the inner Seychelles through the 1998 bleaching event. Trend line represents a Bayesian normal linear-model fit, with the posterior 95% credible and predictive intervals represented as dark and light grey shading respectively.

Figure 3

Relationship between vulnerability of coral reef fish species to climate change disturbance (i.e. bottom-up habitat degradation) and fisheries (i.e. top-down exploitation). The nonlinear negative relationship represents stress-induced community sensitivity, whereby the species pool is greatly reduced if stressors co-occur. The size of the bubbles is proportional to extinction risk. Trend line represents a Bayesian log-Normal second-order polynomial fit with the 95% credible interval represented as grey shading. Blue shading represents hypothetical stress levels.